< Back to previous page

Project

Provoking Diversity in Media using the Diversity Checker

This thesis describes the genesis of the Diversity Checker, a tool meant to provoke journalists into producing more diverse content in an intelligent fashion. Diversity in media is a very broad topic, and the project that this thesis is a part of is similarly broad in its scope. The multidisciplinary nature of the project enables the various disciplines to influence each other positively, and in particular it gives the work presented here a solid grounding in theoretical insights from text mining, corpus and computational linguistics, as well as other relevant fields such as media studies. The work performed in this thesis focuses specifically on the computer science side of diversity, operationalizing and implementing ideas arrived at in cooperation with the other project members. We start out with a thorough review of the literature on automatic text analysis to produce an overview of what is possible. We find in particular that there is a lot of existing research that focuses on English language text, but relatively little that focuses on other languages and as a result, there is a paucity of good data and corpuses available that would suit our purposes. We apply some selected methods from the background research mentioned above to a data set of Flemish news articles that we augment ourselves in a workshop on viewpoint, bias and framing detection. We also gather requirements and based on them implement a first prototype of the Diversity Checker. This prototype is presented to a panel of journalists, and based on their feedback a next version is developed. In the meantime we continue our research on computational enhancement of diversity, which is reported on in papers that form chapters in this thesis but also incorporated into the Diversity Checker. Ultimately we have another round of review with journalists and evaluate how they experience the tool developed, what works well and what could be improved.

Date:1 Sep 2017 →  21 Jul 2019
Keywords:text mining
Disciplines:Applied mathematics in specific fields, Computer architecture and networks, Distributed computing, Information sciences, Information systems, Programming languages, Scientific computing, Theoretical computer science, Visual computing, Other information and computing sciences
Project type:PhD project